Refactor code to remove structural, quality, and efficiency smells without changing behavior.
npx skills add https://github.com/heliohq/ship --skill refactor-heliohq
Please help me install this Agent Skill. Skill: refactor Source: https://github.com/heliohq/ship/tree/main/skills/refactor Command: npx skills add https://github.com/heliohq/ship --skill refactor-heliohq
This Skill improves existing code by removing smells, simplifying logic, and reducing duplication while preserving behavior and keeping verification front and center.
Refactor the selected code path while preserving behavior, then run the repository's tests to confirm the result.
High-intent search queries and answers about installing and using this skill.
To refactor code without behavior drift, you must apply behavior-preserving edits in small batches, simplify functions, and verify the results with tests before and after each change. This ensures structural cleanup maintains predictable execution.
The best way to clean up dead code and simplify long methods is through evidence-based scanning and incremental edits. Breaking up logic and deleting unused state in small batches keeps production repositories stable and maintainable.
Yes, you can consolidate cross-file duplication and replace hand-rolled helpers with existing utilities by refactoring in small batches. Running tests before and after each change batch ensures behavior preservation and prevents test breakage.
Behavior-preserving refactoring handles redundant state and unnecessary comments by normalizing naming conventions and removing obsolete inline documentation. It executes these quality fixes incrementally, validating each step with repository tests to prevent drift.
You should avoid refactoring code in a production repository when you lack sufficient test coverage for evidence-based scanning. Without tests to verify behavior before and after each change batch, structural edits risk introducing unpredictable behavior drift.
Yes, the refactor approach works for both local file edits and broader module cleanup. It applies evidence-based scanning to resolve structural, quality, and efficiency smells across varying scopes while requiring test verification to maintain behavior.
mattphanmRefactor code safely with baseline tests and incremental validation.
gracefullightConvert and deploy TensorFlow models using SavedModel and TensorFlow Lite.
baphledGuide static code analysis tools to detect bugs, vulnerabilities, and style violations.
jmdueaIdentify and resolve performance bottlenecks in Rust applications.
